1. Field of the Invention
Apparatuses and methods consistent with exemplary embodiments according to the present inventive concept relate to video stabilization technology for compensating for movement and shaking of a camera.
2. Description of the Related Art
Conventional video stabilization technology is classified into a video stabilization scheme based on 2-dimensional (2D) analysis and a scheme based on 3-dimensional (3D) analysis.
The video stabilization scheme based on 2D analysis includes estimating a motion, setting a camera motion path through the estimated motion, and generating an image according to the camera motion path. In the estimating of the motion, the motion is measured by analyzing a captured image with a 2D motion model. Although an image is actually obtained by projecting a 3D space into a 2D image in a camera, a 3D motion cannot be correctly modeled with the 2D motion model. Thus, although the video stabilization scheme based on 2D analysis cannot perfectly perform video stabilization, the video stabilization scheme based on 2D analysis can derive an effective result with a relatively small amount of computation.
The video stabilization scheme based on 3D analysis generally reconstructs 3D space information and camera position information from a 2D image by using motion structure analysis technology. A new camera position is defined by using the reconstructed information, and image information regarding the new camera position is synthesized. Although the video stabilization scheme based on 3D analysis requires a considerably greater amount of computation than the video stabilization scheme based on 2D analysis, the video stabilization scheme based on 3D analysis can derive correct video stabilization. However, an image upon which video stabilization has been performed based on 3D analysis may have a problem of image distortion in a partial area of the image.